file:// and content type sniffing

[Axel Hecht]

Hi,
I was just running into a problem with files without extension having no 
content type. The doc-loader sniffs for the mimetype, so once you look 
at the file in the browser, it comes up ok.
But many other paths, like the XML parser, don't and at least the parser 
even locks up.
Stepping thru the code, I saw a nice little point inside 
nsExternalHelperAppService,
http://lxr.mozilla.org/seamonkey/source/uriloader/exthandler/nsExternalHelperAppService.cpp#2267
where one could insert content sniffing for file://.

Opinions?

Axel